The aim was to determine power of nondental background factors to predict behavior-management problems at the first dental visits of 3-yr-old children. A total of 273 children from three kinds of residential area--city, town, and rural area--in Sweden took part. The parents were interviewed before the child's dental visit. The behavior of the child was rated by registering the degree of acceptance according to the method of HOLST & CROSSNER. The following steps were rated: entering the dental treatment room, mirror in mouth, probe on fingernail and tooth, air-blower on hand and in mouth, sitting in the dental chair, and examination. The behavior was then analyzed in relation to the answers of the interviews, and a logistic regression model was used to calculate the power of the variables, separately or combined, to predict behavior-management problems. Seventy-six percent of the children cooperated well at the dental examination; 13% reacted reluctantly, and 11% reacted negatively. Two interview variables had statistically significant predictive power: the parent's expectation of a negative reaction from the child in the dental situation and the child's anxiety when meeting unfamiliar people. Sixty-nine percent of child patients accepted the examination while sitting alone in the chair. Sitting alone used as a predictor for cooperation showed sensitivity 0.80, specificity 0.71, predictive value for positive test 0.44, and predictive value for negative test 0.94.